Nomura's three-year profit forecast for Samsung, 1,800 trillion won, tops its market cap by 131 trillion

Nomura has raised its operating profit forecasts for Samsung Electronics to 391 trillion, 635 trillion and 774 trillion won across three years. The three add up to 1,800 trillion won, 131 trillion more than the 1,669 trillion won market cap at the July 31 close. The same note puts free cash flow at 296 trillion, 456 trillion and 563 trillion won, with shareholder return yields of 11.2%, 17.6% and 22.3% attached. The source lists the three years in order without naming them. The first is taken here as this year.

Samsung Electronics re-enters the $1 trillion club (Seoul Economic Daily English, 2026-07-31)

출처: en.sedaily.com

Seoul Economic Daily reports that on July 31 the shares rose more than 20% during the session, back into the 250,000 won range, and the market cap returned above $1 trillion at $1.158 trillion. Until just before that, a run of declines had pushed the market cap below the $1 trillion line.

Where does that yield come from

Free cash flow is the cash left in hand after the money spent on plants and equipment is taken out of what the business earned. A shareholder return yield is the money going back to shareholders through dividends and buybacks, divided by the market cap.

Samsung's chip unit alone earns 88 trillion won, focus turns to a 120 trillion won shareholder return (Seoul Economic Daily English, 2026-07-30)

출처: en.sedaily.com

Seoul Economic Daily reports that a 120 trillion won shareholder return covering 2027 through 2029 is under discussion, built on the existing policy of returning 50% of free cash flow. So the climb from 11.2% to 22.3% is not a number Nomura imagined on its own. Put the cash flow forecast into a payout policy the company has already stated and the yield falls out mechanically. Which also means the yield is wrong the moment the cash flow forecast is wrong.

Why anyone can write down a number three years out

Memory used to be an industry where even next quarter's price was hard to call. Prices rose, everyone added capacity, and that capacity came back as a glut a few years later. What changed is the contract. On its Q2 earnings call Samsung said it is locking 60% to 70% of memory capacity into five-year long-term agreements. When two thirds of the volume is priced in advance, profit three years out becomes something you can write down. That is why a profit forecast above the market cap does not read only as evidence of undervaluation. Whether the contracts actually break the cycle is still unconfirmed, and the market appears to be holding that unconfirmed part out of the price.

So what settles it

The question comes down to one thing. Whether the 131 trillion won between the market cap and the forecast is the market's mistake, or a fair discount on the forecast. The first year is what settles it. If second-half operating profit clears four times the Q2 run rate and lands near 391 trillion won, the remaining two years carry the same weight. If it does not, the 635 trillion and 774 trillion figures were already off in year one. The price of a memory company has always been set by how early the next downturn is pulled into it. If the long-term agreements really do delay that downturn, today's discount loses its footing, and if they do not, the forecast is what breaks. A market cap that slipped under $1 trillion within a month and then recovered more than 20% in a single day says the market is swinging on this one assumption.